INDEX

Explanations

words related to video game elements and interactions

oai_token-act-pair · gpt-3.5-turbo

elements related to interactive gaming experiences

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.hook_resid_pre

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

torch.float32

Hook Point

blocks.10.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ļéĨĴ

-0.85

quickShipAvailable

-0.84

soType

-0.81

inventoryQuantity

-0.75

emporary

-0.75

statement

-0.74

endif

-0.70

paren

-0.70

 Disclosure

-0.68

POSITIVE LOGITS

 quests

1.18

 missions

1.13

 turrets

1.08

 traverse

1.06

 objectives

1.03

 enemies

1.01

 unlock

0.99

 NPCs

0.99

 travers

0.98

 mech

0.98

Activations Density 0.493%

words related to video game elements and interactions

elements related to interactive gaming experiences

No Comments

No Known Activations